From c04f91f98a2465d4d0319aa83fb72dfbae40e22f Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:00:29 +0000 Subject: [PATCH 01/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh index 55f195dd2..8aa7e0f78 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh @@ -5,7 +5,7 @@ cur_path=`pwd` #export ASCEND_SLOG_PRINT_TO_STDOUT=1 #export GE_USE_STATIC_MEMORY=1 -export HCCL_CONNECT_TIMEOUT=300 +export HCCL_CONNECT_TIMEOUT=1200 #集合通信参数,不需要修改 export RANK_SIZE=8 -- Gitee From c8af0d52c81d32982a1e9ae726e9c85cbaa7515e Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:20:13 +0000 Subject: [PATCH 02/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh index be3561702..003749e50 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh @@ -5,7 +5,7 @@ cur_path=`pwd` #export ASCEND_SLOG_PRINT_TO_STDOUT=1 #export GE_USE_STATIC_MEMORY=1 -export HCCL_CONNECT_TIMEOUT=300 +export HCCL_CONNECT_TIMEOUT=1200 #集合通信参数,不需要修改 export RANK_SIZE=4 -- Gitee From d4e3c4296ce05372ddf1430df3b6993665bd4a8a Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:44:53 +0000 Subject: [PATCH 03/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh. --- .../test/train_performance_4p.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh index 003749e50..9a842fcee 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh @@ -24,8 +24,8 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -train_size=13107200 -display_step=1 +train_size=13107200 * 2 +display_step=40 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From 4d76123f291786ce4e670042b0461420d45aeba3 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:45:48 +0000 Subject: [PATCH 04/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh. --- .../test/train_performance_8p.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh index 8aa7e0f78..2905fe99f 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh @@ -24,8 +24,8 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -train_size=13107200 -display_step=1 +train_size=52428800 +display_step=80 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From cefb1daad47133ab8cc2f3baf59830cea53bb411 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:48:19 +0000 Subject: [PATCH 05/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh index 9a842fcee..79bb17058 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh @@ -24,7 +24,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -train_size=13107200 * 2 +train_size=52428800 display_step=40 n_epoches=4 -- Gitee From e1d42466d48c603af785bdc04d12745c023d0aff Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:50:34 +0000 Subject: [PATCH 06/15] update WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh index 5b740ba40..176b3f14a 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh @@ -26,7 +26,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -display_step=10 +display_step=40 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From 6acc4729177f78086a05c40a79ef08ea99aac046 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:50:52 +0000 Subject: [PATCH 07/15] update WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh index 953c6a895..2c4824001 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh @@ -26,7 +26,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -display_step=10 +display_step=80 n_epoches=8 #维持参数,以下不需要修改 -- Gitee From 39b9afa2c84531894dd46b972c6e99a1b052ff9e Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 09:59:01 +0000 Subject: [PATCH 08/15] update WideDeep_ID2712_for_TensorFlow/train.py. --- .../recommendation/WideDeep_ID2712_for_TensorFlow/train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py index 79b8896a0..049e0462d 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py @@ -458,7 +458,7 @@ if __name__ == '__main__': dt = end_time_test - start_time_test fps=train_para['batch_size'] * rank_size * config.iterations_per_loop/dt - + print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) if epoch_finished_batches % (display_step / num_gpu) == 0: # print step if _epoch: print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) -- Gitee From a3bf35be5b1cd112f1f969d5b30e21c70145c0fb Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:10:15 +0000 Subject: [PATCH 09/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh index 2905fe99f..bb0e0a844 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_8p.sh @@ -25,7 +25,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" train_size=52428800 -display_step=80 +display_step=10 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From 9b085e7bc649dd1ab5a3b9f0867d0e1ce799581e Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:11:18 +0000 Subject: [PATCH 10/15] update WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh index 79bb17058..829647822 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_performance_4p.sh @@ -25,7 +25,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" train_size=52428800 -display_step=40 +display_step=10 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From 4e3e1051c89c1f7066650ee1052a5ca03c40845e Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:11:36 +0000 Subject: [PATCH 11/15] update WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh index 2c4824001..953c6a895 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_8p.sh @@ -26,7 +26,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -display_step=80 +display_step=10 n_epoches=8 #维持参数,以下不需要修改 -- Gitee From 5da98e6cd78ad1b47c0916d6cd67c9ddc02b99ad Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:11:56 +0000 Subject: [PATCH 12/15] update WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh. --- .../WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh index 176b3f14a..5b740ba40 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/test/train_full_4p.sh @@ -26,7 +26,7 @@ RankSize=1 #参数配置 data_path="/npu/traindata/ID2940_CarPeting_TF_WideDeep_TF" -display_step=40 +display_step=10 n_epoches=4 #维持参数,以下不需要修改 -- Gitee From 9e0ac1bedcb12ed48d7f338716a8d751babffb71 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:22:51 +0000 Subject: [PATCH 13/15] update TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py. --- .../recommendation/WideDeep_ID2712_for_TensorFlow/train.py | 1 - 1 file changed, 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py index 049e0462d..3282846ad 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py @@ -458,7 +458,6 @@ if __name__ == '__main__': dt = end_time_test - start_time_test fps=train_para['batch_size'] * rank_size * config.iterations_per_loop/dt - print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) if epoch_finished_batches % (display_step / num_gpu) == 0: # print step if _epoch: print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) -- Gitee From 410ee4a407db86fd0c8362896b75f294e1921654 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:23:28 +0000 Subject: [PATCH 14/15] update TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py. --- .../recommendation/WideDeep_ID2712_for_TensorFlow/train.py | 1 + 1 file changed, 1 insertion(+) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py index 3282846ad..e89ef65ea 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py @@ -458,6 +458,7 @@ if __name__ == '__main__': dt = end_time_test - start_time_test fps=train_para['batch_size'] * rank_size * config.iterations_per_loop/dt + if epoch_finished_batches % (display_step / num_gpu) == 0: # print step if _epoch: print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) -- Gitee From 53621883b14aeb20794df362ced1db88cfca80b8 Mon Sep 17 00:00:00 2001 From: majun121 <867479212@qq.com> Date: Wed, 20 Apr 2022 10:24:20 +0000 Subject: [PATCH 15/15] update TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py. --- .../recommendation/WideDeep_ID2712_for_TensorFlow/train.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py index e89ef65ea..79b8896a0 100644 --- a/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py +++ b/TensorFlow/built-in/recommendation/WideDeep_ID2712_for_TensorFlow/train.py @@ -458,7 +458,7 @@ if __name__ == '__main__': dt = end_time_test - start_time_test fps=train_para['batch_size'] * rank_size * config.iterations_per_loop/dt - + if epoch_finished_batches % (display_step / num_gpu) == 0: # print step if _epoch: print("================epoch_finished_batches", epoch_finished_batches, display_step, num_gpu) -- Gitee